RemoteVAR：面向遥感变化检测的自回归视觉建模

RemoteVAR: Autoregressive Visual Modeling for Remote Sensing Change Detection

January 17, 2026

作者: Yilmaz Korkmaz, Vishal M. Patel

cs.AI

摘要

遥感变化检测旨在定位并表征两个时间点之间的场景变化，是环境监测与灾害评估等应用的核心技术。当前视觉自回归模型虽展现出卓越的图像生成能力，但由于可控性弱、密集预测性能欠佳及曝光偏差等问题，其在像素级判别任务中的应用仍受限。本文提出RemoteVAR——一种基于自回归模型的新型变化检测框架，通过交叉注意力机制将多分辨率融合的双时相特征嵌入自回归预测过程，并采用专为变化图预测设计的自回归训练策略，有效解决了上述局限。在标准变化检测基准上的大量实验表明，RemoteVAR相较于基于扩散模型和Transformer的强基线模型均取得显著提升，为遥感变化检测提供了具有竞争力的自回归解决方案。代码将发布于https://github.com/yilmazkorkmaz1/RemoteVAR。

English

Remote sensing change detection aims to localize and characterize scene changes between two time points and is central to applications such as environmental monitoring and disaster assessment. Meanwhile, visual autoregressive models (VARs) have recently shown impressive image generation capability, but their adoption for pixel-level discriminative tasks remains limited due to weak controllability, suboptimal dense prediction performance and exposure bias. We introduce RemoteVAR, a new VAR-based change detection framework that addresses these limitations by conditioning autoregressive prediction on multi-resolution fused bi-temporal features via cross-attention, and by employing an autoregressive training strategy designed specifically for change map prediction. Extensive experiments on standard change detection benchmarks show that RemoteVAR delivers consistent and significant improvements over strong diffusion-based and transformer-based baselines, establishing a competitive autoregressive alternative for remote sensing change detection. Code will be available https://github.com/yilmazkorkmaz1/RemoteVAR{here}.